Synthesis of 9, 10-cyclobutenophenanthrene from 9, 10-dimethylene-9

Bioethics détente. In 2015, Jennifer Doudna, codeveloper of the CRISPR/Cas9 gene-editing technology, convened a meeting... SCIENCE CONCENTRATES ...
0 downloads 0 Views 781KB Size
J. Org. Chem., Vol. 37, No. 7 , 1972

9, 10-CYCLOGUTENOPHENANTHRENE However, three of the four coupling constants differ substantially from those reported previously.2 Uv had A 258 mM shoulder, 265.5 (e 1026), 272.5 (1036). cis-9,10-Diisopropyl-O,lO-DHA.-To g-isopropy1-9,lO-DHA (1.26 g, 5.6 mmol) in dry T H F (30 ml) at -60" was added nbutyllithium (6.0 mmol). The reaction mixture was stirred a t 0' for 1 hr and the reaction was terminated by the rapid addition of excess isopropyl iodide. After separation of salts with water and removal of ether solvents, an nmr spectrum indicated the presence of 12.5y0 cis- and 87.5% trans-9,10-diisopropyl-9,10DHA by integration of the benzylic hydrogen doublets at 6 3.78 (J = 5.0 Ha for the trans isomer) and 3.27 (J = 9.5 Hz for the cis compound). See earlier l i t e r a t ~ r e . ~ After chromatography over dry basic alumina (hexane), 0.85 g of trans-9,10-diisopropyl-9,10-DHA (57%), mp 73-74', was obtained after recrystallization from ethanol (lit.3 mp 76-77'); uv 257 mp (shoulder), 266 (E 647), 272 (588). In later fractions the cis isomer appeared predominantly as an oil which crystallized upon trituration with ethanol to yield 60 mg of cis-9,lO-diisopropyl-9,1O-DHA (4y0, mp 99.5-105'), lit.3 mp 109-110'; uv h 258 mp ( e 822), 263 (1084), 272 (1221). Methyllithium Addition to Anthracene.-Anthracene (0.5 g, 2.8 mmol) in 50 ml of T H F was mixed with excess methyllithium (14 mmol) in ether. After refluxing for 4 hr, excess isopropyl iodide was added quickly. After 1 hr stirring, salts were separated with water and gas chromatography showed the presence of five components. The ratio of cis- and trans-9-isopropyl-10methyl-9,lO-DHA was 40:60 =t 1 as determined by vpc. Unchanged anthracene was recovered. trans-9-Ethyl- 10-methyl-9,lO-DHA. 16-Lithium (0.15 g )-am(15) We thank Mr. Isaac Angres (NSF undergraduate participant) for running this experiment.

1015

monia (300 nd) reduction of 9-ethyl-10-methylanthracene(2 g ) in T H F (120 ml) for 3.5 hr was followed by addition of ethanol (10 ml) and HzO (10 ml). Solvents were evaporated and the oil obtained from ether-water treatment was recrystallized from absolute ethanol to give 1.3 g of white needles (65oj,), mp 33-34'. Anal. Calcd for C17H18: C, 91.84; €1, 8.16. Found: C, 91.73; H,8.25; C,91.75; H,8.38. Uv had h 212.2 mp (e 19,480), 264.5 (1140), 271.8 (1067); nmr 6 0.887 (t, 3, CHaCH2-, J = 7.0 Hz), 1.71 (d, 3, CH~CIOH, J = 6.7 Hz), 3.80 (t, 1, CsHCH2-, J = 7.5 Hz), 3.99 (9, 1, CloHCH3,J = 6.7 Hz), 7.22 (m, 8, aromatic). These chemical shifts and coupling constants do not agree well with those published previ~usly.~This sample was purified by recrystallization before nmr spectroscopy and spectra were obtained on an HA-100 instrument better suited for careful determination of coupling constants. There is no doubt, however, of the identity of this material with that described previously.6

Registry No.--cis-2, 21438-93-7 ; trans-2, 33608-27-4 ; 9-isopropy1-9,10-dihydroanthracenel17573-50-1; trans9,10-diisopropyl-9,lO-dihydroanthracene,25340-82-3; cis-9, l0-diisopropyl-9,10-dihydroanthracene12431621-0; trans-9-ethyl-lO-methyl-9,1O-dihydroanthracene, 23660-35-7. Acknowledgment. -This investigation was supported in part by Undergraduate Science Education Program Grants CY-5723 and GY-7540, National Science Foundation.

The Synthesis of 9,lO-Cyclobutenophenanthrenefrom 9,10-Dirnethylene-9,1O-dihydrophenanthrene JOHN P. ANHALT,*EARLW. FRIEND, JR.,AND EMILH. WHITE^ Department of Chemistry, The Johns Hopkins University, Baltimore, Maryland 11118 Received May 18, 1971

9,10-Dimethylene-9,10-dihydrophenanthrene (4)has been prepared from trimethyl( 10-methyl-9-phenanthrylmethy1)ammonium chloride ( 6 ) and characterized by its physical data and the formation of an adduct with maleic anhydride. Irradiation of compound 4 yielded 9,lO-cyclobutenophenanthrene(2), the proof of structure of which is discussed.

Our interest in 9,lO-cyclobutenophenanthrene( 2 ) was first aroused during a study of the photochemistry of 1,2-diphenylcyclobutene (l),in which it is a possible product (eq 1).2 We were stimulated to the synthesis 3

of 2 by the subsequent report by Masamune and Kat0 of diphenyltetrahedrane (3).3 This report drew our attention because of our interest in 3 and because the physical properties attributed to 3 by Masamune and Kat0 appeared to match better the predicted proper(1) Author to whom inquires should be addressed. (2) E. H. White and J. P. Anhalt, Tetrahedron Lett., 3927 (1965). (3) (a) S. Masamune and M. Kato, J . Amer. Chem. Soc., 8'7, 4190 (1965): (b) 9. Masamune and M. Kato, ibid., 88, 610 (1966).

4

ties of 2.4 We wish now to report the full synthesis and characterization of 2. During the progress of this synthesis, we were able to demonstrate the intermediacy of 9,lO-dimethylene9,lO-dihydrophenanthrene(4). Previously, this compound had been reported as a reactive intermediate and its presence was inferred only by trapping with various dien~philes.~The instability of 4 with respect to dimerization and polymerization prevented our complete characterization of it; however, we were able to obtain its ultraviolet spectrum in dilute solution. The direct observation of 4 in the ultraviolet is to our knowledge the first such observation of an o-quinodi(4) E. H.White, G. E. Maier, R. Graeve, U. Zirngibl, and E. W. Friend, i b i d . , 88, 611 (1966). (5) (a) I. T.Millar and K. V. Wilson, J . Chem. SOC.,2121 (1964): (b) J. K. Stille and R. T. Foster, J . Org. Chem., 28, 2708 (1963); ( 0 ) P. D. Gardner and H. S. Sarrafizadeh R., J . Amer. Chem. Soc., 82,4287 (1960).

1016 J. Org. Chem., Vol. 37, No. 7, 2972

ANHALT,FRIEND, AND WHITE

methane not substituted in the terminal methylene positions. 9,lO-Cyclobutenophenanthrene (2). -The synthesis of 2 was achieved in five steps starting with the readily available 9,lO-phenanthroquinone (5). The conversion of 5 to trimethyl(l0-methyl-9-phenanthrylmethy1)ammonium chloride (6) was effected by known proc e d u r e ~ . ~This intermediate was purified as the monohydrate. The conversion of the monohydrate of 6 to cyclobutenophenanthrene (2) was straightforward (eq 2). O w 0

wCH2&(CH&1-

tert.BuOK lert.BuOH



4

hv 0-

2537 A

2

(2)

6

It was found that best yields are obtained when dry tert-butyl alcohol is used, and when both steps in the reaction are carefully monitored (see Experimental Section). Early irradiation of 6 before elimination is completed appeared to give more complicated reaction mixtures. I n addition, compound 2 is photodecomposed by prolonged irradiation. The physical data found for the product are completely consistent with the assigned structure of 2. In particular (see Experimental Section for other data) the ultraviolet spectrum of 2 [h$Yane 255 mp (log B 4.83) 3 is very similar to that of 9,lO-dimethylphenanthreneand clearly indicates a phenanthrene chromophore. The mass spectrum of 2 shows apparent successive losses of 15, 13, 13, and 13 mass units. The pattern of fragmentation may be explained by two parallel processes (eq 3).

-

-e-

M.+ m/e 204

Benzylic protons, 6

o-Xylenea 2.23 Tetralin” 2.76 Indana 2.91 Benzocyclobuteneblc 3.14 9,lO-Dimethylphenanthreneo 2.65 Cyclobutenophenanthrenec 3.35 a N. S. Bhacca, D. P. Hollis, L. F. Johnson, and E. A. Pier, “NMR S,pectra Catalogue,” Vol. 11, Varian Associates, Palo Alto, Calif., 1963. I n chloroform-&. G.Fraenkel, Y. Asahi, M. J. Mitchell, and M. P. Cava, Tetrahedron, 20, 1179 (1964). 0 I n carbon tetrachloride.

Furthermore, the increased downfield shifts of the 1,8 protons in 9,lO-substituted phenanthrenes7&(van der Waals d e ~ h i e l d i n gare ) ~ ~not observed in 2 (Table 11),a

H3C

2

TABLE I POSITION OF BENZYLIC PROTONS AS A FUNCTION OF RINGSIZEIN BENZOCYCLOALKENES

TABLE I1 POSITION OF RINQPROTONS AS A FUNCTION OF SUBSTITUTION IN PHENANTHRENES~ --Ring protons, ab-Phenanthrene

2,3,6,7

18

4!5

7.51

7.74

8.56

9,lO-Dideuterio-2,7-dimethylphenanthrenec 7.3 7.6 8.4 9,lO-Cyclobutenophenanthrene(2) 7.45 7.65 8.60 9,lO-Dimethylphenanthrene 7.47 7.98 8.57 9-Methoxymethyl-10-methylphenanthrene 7.61 8.15 8.67 a Spectra, except where noted, were obtained in carbon tetraCenchloride using a Varian Associates HA-100 spectrometer. ters of multiplets in most cases. Positional assignments are made in accord with assignments given in ref 7 for phenanthrenes. c L. A. Paquette, J. Amer. Chem. Soc., 86,4085(1964);no conditions given.

result, presumably, of the smaller size of the cyclobutane ring relative to two methyls. 9,10-Dimethylene-9,1O-dihydrophenanthrene (4).The intermediacy of quinodimethane 4 in the synthesis of compound 2 (eq 2) was demonstrated by trapping it with maleic anhydride (in experiments without irradiation), The nearly quantitative yield of 1,2,3,4-tetrahydro-2,3-triphenylene-cis-dicarboxylic acid (’7)6b isolated after hydrolysis indicates that almost complete conversion of 6 to 4 had occurred (Table 111). The small amount of dimer found may have formed after addition of maleic anhydride. The formation of 8

m/e 163

7

8

The nmr spectrum is consistent in detail with structure 2. The methylenic protons of 2 a t 6 3.35 are shifted 0.70 ppm downfield from the methyl protons of 9,lO-dimethylphenanthrene; this shift is consistent with the presence of a fused cyclobutane ring (Table I).

was not evident in ultraviolet spectra of reaction mixtures immediately after being quenched with acetic acid; however, it is not certain that such a small quantity would be detected. When the addition of maleic anhydride to 4 was delayed, dimerization of the

(6) G. Quinkert, M . Finke, J. Palmowski, and W-W ’Wiersdorff, Mol. Photochem., 1, 433 (1969), and G. Quinkert, Photochem. Photobiol., ‘I, 783 (I96S), have observed diphenyl- and tetraphenyl-o-xylylene a t low temperature.

(7) (a) P. M . Bavin, K. D. Bartle, and J. A. 6 . Smith, Tetrahedron, 21, 1087 (1965); (b) N. S. Bhacca and D. H. Williams, “Applications of Nmr Spectroscopy in Organic Chemistry,” Holden-Day, San Francisco, Calif., 1964, p 183.

J. Org. Chem., Vol. 37, No. 7 , 1976

9, 10-CYCLOBUTENOPHENANTHRENE

PRODUCT YIELDSFOR Aliquot

a

THE

1017

TABLE I11 REACTION OF ~,~~-DIMETHYLENE-~,~~-DIHYDROPHENANTHRENE (4) WITH MALEICANHYDRIDE

Time of addn of maleic anhydride, hra

----mg

29.0 0.1 Ai 29.0 Aa 0.1 B 8.0 12.3 2.8 C 103.0 After ultraviolet spectra indicated complete conversion to 4.

Compd 7-mmol

%b

91.0 0.091 93.0 0.093 38.0 0.038 8.7 0.0087 Based on 0.100 mmol of 6.

,------Compd mg

mmol

%b

0.7 0.5 10.3 17.1

0.0017 0.0012 0.025 0.042

3.4 2.4 50.0 84.0

--8 -

TABLE IV ULTRAVIOLET SPECTRA O F MODEL2,3-DIPHENYL-1,3-DIENES Xmax, mfi (log -----)6

a

216 sh (4.56), 244 (4.42), 260 sh (4.25), 300 (3.79)" 9,lO-Dimethylene-9,1O-dihydrophenanthrene(4) 243 (4,26), 280 sh (3.22), 287 sh (2.70)b 2,3-Diphenylbutadiene" 247 (4.41), 295 sh (3.00)b 2,3-Diphenyl- 1,3-cyclo~ctadiene~ In tert-butyl alcohol. b In cyclohexane. c A. C. Cope and D. S. Smith, J.Amer. Chem. Soe., 74,5136 (1952).

unstable 4 did occur to yield compound 8. The yields of 7 and 8 as a function of time (Table 111)indicate that quinodimethane 4 has a half-life of about 8 hr at room temperature at a concentration of M in tert-butyl alcohol. Comparison of the ultraviolet spectrum of the quinodimethane with model systems (Table IV) indicates that the spectrum is consistent with structure 4. Since the quinodimethane 4 could not be isolated, concentration of solutions of 4 leading to dimer 8 and polymer, the e values we report are approximate and are calculated by assuming the concentration of 4 t o be the same as the initial concentration of 6. This assumption is justified since high yields (93%) of the maleic anhydride adduct 7 were obtained (Table 111). Furthermore, the ultraviolet spectra of solutions of 4 briefly irradiated with a 4-W germicidal lamp were interpretable as the sum of just the two components 2 and 4 (see Experimental Section). Since the reactant 6 and known products 2 and 8 have greater extinction coefficients at the position of maxima for 4, the presence of small amounts of these compounds would lead to an increase in the calculated E values for 4. The absence of these compounds in the reaction mixture after apparent complete conversion of 6 to 4 is evident from the constant e values obtained for 4 over a range in concentrations from 2.0 X to 5.3 X 10-4

M. Each of the model compounds in Table IV shows some evidence of long wavelength absorption between 280 and 300 mp. The presence of a well-defined maximum at 300 mp in 4 may be the expression of a relatively rigid structure compared to the model compounds. Experimental Section Melting points, except where noted, were taken with a ThomasHoover capillary melting point apparatus and are uncorrected. Joseph Walters Elemental analyses were preformed by either Mi-. in this department or Galbraith Laboratories, Inc., Knoxville, Tenn. Infrared spectra were determined on Perkin-Elmer Model 337 or 521 infrared spectrometers and were calibrated against known absorption bands of polystyrene. Ultraviolet spectra were determined on a Gary Model 14 spectrometer. Proton magnetic resonance spectra (nmr) were determined on Varian Associates A-60 or HA-100 spectrometers using tetramethylsilane as an internal standard for nonaqueous solutions and sodium 2,Z-dimethyl-2-silapentane-5-sulfonate (DSS) for aqueous solutions.

Mass spectra were determined on a Hitachi Perkin-Elmer RMU-6 mass spectrometer. Thin lager chromatography (tlc) was performed on Eastman chromatogram sheets eo-ntaining a fluorescent indicator. Visualization was with 2537-8 light. 9-Chloromethyl-l O-methylphenanthrene.-9-Chloromethyl-10methylphenanthrene was prepared by a modification of the method of Millar and W'ilson.5a To a suspension of 9,lO-dimethy1-9,lO-dihydroxy-9, 10-dihydrophenanthrenes (10.58 g, 44.0 mmol) in ether (60 ml), thionyl chloride (12.7 ml, 21.0 g, 0.18 mol) was added with stirring. The reaction mixture was protected from atmospheric moieture with a drying tube. After stirring for 1.5 hr the mixture was heated to reflux for 5.5 hr at which time the reaction mixture became homogeneous. Reflux was allowed to continue for 2 hr, after which time the ether was allowed to distill off until stirring became impossible and hydrogen chloride evolution commenced. The mixture was then placed under aspirator vacuum and heated on a steam bath until hydrogen chloride evolution ceased ( I hr). Ethyl acetate (100 ml) was added and the mixture was heated to reflux overnight. The nearly homogeneous hot solution was filtered and allowed to cool to give 4.25 g of 9-chloroniethyl-10-methylphenanthrene(17.7 mmol, 407,): mp 132.5-154" (lit.5~155-156'); ir (KBr) 1254, 780, 760, and 720 cm-1. The ethyl acetate mother liquor was evaporated to dryness on a rotary evaporator and the residue was dissolved in hot benzene. Addition of an equal volume of isooctane gave nicely shaped, slightly colored crystals of 9-chloromethyl-10-methylphenanthrene (1.40 g, 5.8 mmol, 13%), mp 146-148.5'. Tlc of each batch on silica gel (benzene-hexane, 1:2) gave a single spot of 0.48. The benzene-isooctane mother liquor on evaporation gave a crystalline mass (4.52 g) which on tlc on silica gel (benzenehexane, 1:2) gave two spots of Rf 0.27 and 0.48. This material was chromatographed on silica gel (100 g) using a benzene-hexane mixture (1:2) as eluent. The fractions containing the slow moving material were evaporated t o crystalline solids and combined. Itecrystallization from hexane gave 2.45 g (10.6 mmol, 24%) of lO,lO-dimethyl-9( 10H)-phenanthrone (i), mp 72.0-73.5" (lit.e 75').1° The infrared spectrum (KBr) was identical with that reported in the literature for i12 with prominent bands at (8) 9,lO-Dimethyl-9,10-dihydroxy-Q,lO-dihydrophenanthrenewas prepared b y the addition of 9,10-phenanthraquinone t o a solution of methylmagnesium iodide in ether according to the method of Gardner and Sarrafizadeh." The product after recrystallization from benzene melted a t 162.5-163.8' (lit." 163-164°). (9) T . Zinckeand W. Tropp, JustusLiebigs Ann. Chem., 363, 242 (1908). (10) The production of i had not previously been reported in the synthesis of 9-chloromethyl-10-methylphenanthrene b y this method.11 The occurrence here may have been due to the somewhat more vigorous conditions employed.

i (11) 9. Hauptmann, Chem. Ber., 98, 2604 (1960). (12) A. Sohonberg and G . Schutz, ibid., 95, 2386 (1962).

1018

J. Org. Chem., VoZ. 37, No. 7,1972

1675, 982, 783, 756, and 732 cm-l; uv (hexane) 238 mp (log B 4.39), 248 (4.33), 267 (3.99), 276 (4.05), 292 sh (3.74), and 320 (3.50); nmr (CClr) S 7.92 (m, 3.2 H), 7.38 (m, 5.2 H), and 1.50 (s, 6.0 H). The fraction containing the faster moving material gave a crystalline solid after evaporation. Tlc on silica gel (benzene-hexane, 1:2) of this material was identical with that of 9-chloromet hyl- 10-methylphenanthrene. Trimethyl( 10-methyl-9-phenanthrylmethy1)ammoniumChloride Monohydrate (6).-The procedure of Millar and Wilson68 for the synthesis of trimethyl(l0-methyl-9-phenanthrylmethyl)ammonium chloride (6) was followed. Complications attend the synthesis, and thus our procedure is described in detail. Trimethylamine generated by stirring the corresponding hydrochloride (9.6 g, 0.10 mol) with barium oxide (26 g, 0.27 mol) was bubbled into a stirred suspension of 9-chloromethyl-10-methylphenanthrene (3.000 g, 12.45 mmol) in a mixture of chloroform (210 ml) and absolute methanol (70 ml). After stirring for 1.5 hr a t room temperature, the reaction mixture was heated to reflux for 1.0 hr at which time the solution was homogeneous. The reaction mixture was allowed to cool with stirring for an additional 3 hr. Tlc on silica gel (benzene-hexane, 1:2) showed the presence of considerable starting material. Additional trimethylamine generated by adding dropwise a solution of the hydrochloride (14.4 g, 0.150 mol) in water (50 ml) to a huge excess of sodium hydroxide (200 g) was bubbled through the reaction mixture for 1.0 hr a t room temperature. Tlc under the same conditions no longer showed starting material; however, a new spot at Rf0.18 was present in addition to the expected spot at the origin corresponding to the title quarternary chloride. The reaction mixture was then heated to reflux for 2.0 hr, after which time the tlc was unchanged. The solvents were removed from the reaction mixture on a rotary evaporator and the resulting solid was placed under high vacuum overnight. The solid was then dissolved in absolute ethanol and the solution was filtered. Addition of ether to the filtrate gave 1.958 g of material which gave only a very faint spot of Rf 0.20 in addition to the spot a t the origin. The nmr (CDC18)of this material showed a doublet at 8 2.93 (J = 5 Hz) attributable to trimethylamine hydrochloride.18 The only other peaks in the spectrum were attributable to the quaternary chloride 6 at 6 8.72 (m, 3.0 H), 8.17 (m, 1.2 H), 7.70 (m, 3.7 H), 5.82 [broad doublet (unresolved AB quartet), 1.7 H, CHz], 3.50 (s, 9.0 H), and 3.02 (s, 3.6 H). Integration of this spectrum indicated 9.0% (31 mol %) of trimethylamine hydrochloride was present. When the spectrum was taken using deuterium oxide as solvent the doublet attributed to trimethylamine hydrochloride coalesced to a singlet at 6 3.17,14 which also indicated the presence of about 9% of trimethylamine hydrochloride. The presence of trimethylamine hydrochloride was further confirmed by weak infrared bands (Npjol) a t 2515, 2470, and 987 cm-1. The quaternary chloride 6 readily formed a chloroform insoluble hydrate. The impure quaternary chloride 6 was crystallized twice from water to give 1.103 g (3.47 mmol, 28%) of analytically pure trimethyl(l0-methyl-9-phenanthrylmethy1)ammonium chloride monohydrate (6): uv max (tertbutyl alcohol) 224 mp (log B 3.38), 250 (4.63), 257 (4.69), 272 sh (4,11), 278 sh (3.99), 292 (3.96), 303 (3.93), 334 (2.53), 340 sh (2.26), and 350 (2.26); ir (Nujol) 3430, 3350, 875, and 760 cm-l; nmr (DzO) 8 8.21 (m, 2.0 H), 7.91 (m, 1.9 H), 7.68 (m, 3.8 H), 4.82 [broad doublet (unresolved AB quartet), 1.7 HI, 2.98 (s, 8.9 H), and 2.52 (9, 3.1 H). Anal. Calcd for ClQHznNCl-H20: C, 71.80; H, 7.61; N, 4.41. Found: C, 71.81; H, 7.36; N, 4.22. The mother liquor from the initial precipitation of the quaternary ammonium chloride with ether was evaporated to a crystalline solid (1.638 g) which gave a principal spot on tlc on silica gel (benzene-hexane, 1:2) of Rf 0.20. This material was dissolved in carbon tetrachloride (10 ml) and filtered. The clear solution was then evaporated to give a crystalline solid which was recrystallized from ethanol to give 0.598 g (2.54 mmol, 20%) of 9-methoxymethyl-10-methylphenanthrene: mp 108.7-111.0' (lit.6c 113-114'); ir (KBr) 1099, 953, 755, and 720 cm-l; nmr (CDCI,) S 8.67 (m, 2.2 H), 8.15 (m, 2.2 H), 7.61 (m, 4.0 H), 4.96 (s12.0H),3.48(~,2.8H),and2.76(~,3.0H). 9,10-Dimethylene-9,1O-dihydrophenanthrene(4).-A freshly (13) A nearly saturated solution of trimethylamine hydrochloride in CDCla gave a singlet at 8 2.95; however, a dilute solution gave a doublet a t 8 2.88 (J = 5 Ha). (14) The nmr (DzO) of trimethylamine hydrochloride gave a singlet at 6 2.91. Addition of trimethyl(l0-methyl-9-phenanthrylmethy1)ammonium chloride (